Experts say that the first large-scale job automations are to begin in early 2030s – but have they already begun?
You go to work. You work hard, do your job and collect your wages. For hundreds of years, the average person has known and expected this as one of the fundamental cornerstones of how civilization works. Societies function because – in theory – everybody has to contribute something of value to society in order to reclaim valuable resources such as food and shelter for themselves.
But what if that were to change? What if advances in technology, robotics and Artificial Intelligence (A.I) made humans an unnecessary part of running a human society?
Depending on your viewpoint, this concept is exciting or terrifying. For many, it’s a combination of both.
But regardless of your personal opinions on this matter, job automation – the replacing of human labour with robots and A.I – is a reality we will be facing as major advancements in Mobile Robotics (MR) and Machine Learning (ML), alongside pattern recognition algorithms advancing at a rapid rate mean that robots will soon be inarguably better at our jobs than humans. Recent technological advances such as Google’s driverless cars, virtual assistants such as Siri and Cortana and Japanese healthcare robots are all examples of technology beginning to tackle jobs many thought only humans could handle. So what else can robots do?
Jobs at Risk from Job Automation
Research conducted by PriceWaterhouseCoopers (PWC) – the most recent study conducted into the subject and thus far providing the most moderate results – indicates that around 30% of all jobs are at high risk (70% likelihood or more) from job automation by the early 2030s. (This study is focused on the U.K and all results mentioned are primarily indicative of the U.K)
Across all industries and sectors, there are a few more patterns we can observe to assess which jobs are at risk from job automation in the near future. Generally speaking, it is the low-skill, low-education and low-wage jobs that are the most susceptible to job automation. However, the majority of the jobs at stake are concentrated into four key sectors which together make up 36% of U.K’s employment share. These industries are:
Transportation and Storage, with 56% or 0.95 million jobs at risk of the industries’ jobs at risk.
Manufacturing, with 46% or 1.22 million jobs at risk.
Wholesale and Retail, with 44% or 2.25 million jobs at risk.
Administrative and Support Services, with 37% or 1.09 million jobs at risk.
Why these industries? Well, the primary factor that dictates whether a job is considered high risk or low risk from job automation is the composition of tasks that job requires.
Jobs which consist of many routine and manual tasks are more likely to result in job automation than others. The reasons behind this shouldn’t be too hard to understand; a robot does not tire in the same way a human does and does not legally or physically require breaks or any time off work. From a pure efficiency standpoint, no human can compete with a machine that can work 24 hours a day, 7 days a week, 365 days a year. In addition, the routine aspect of technology is something we’ve come to expect – any well designed piece of hardware or software does exactly what we want it do when given the right instructions. If it doesn’t, we’re likely to complain. Therefore, in addition to the theoretically ceaseless labour a machine can undertake, the machine is also capable of producing the exact same, high-quality results day after day.
All of the four main sectors described above where found to consist of mainly routine and manual tasks, leaving them the most vulnerable to job automation. Although it is a much smaller (Employing 0.6% of people) sector than the previously mentioned ones, the waste and sewage management sectors has the largest percentage of its jobs at risk at 62% due to its heavy reliance on manual and routine tasks, many of which are undesirable for humans.
On the flipside, jobs heavily reliant on social and literacy skills, such as Education and Health and Social Work jobs faced a much lower risk of job automation - 8% and 17% of jobs respectively. Again this should be easy to understand, as we find it much more natural to bond and empathize with other humans as opposed to machines, meaning jobs that involve lots of social interaction are relatively safe from job automation for the time being. Despite some incredible advancements, machines and A.I are also quite a while away from being able to read, write and understand text to the levels that humans are capable of.
Of course, job automation will not only bring job loss, but also job creation, especially in the fields of robotics and A.I. It is estimated that 5% of jobs in 2030 will be ones that do not exist now and while this does not offset the 30% job loss we mentioned earlier, it does show there are jobs to be gained from job automation for those in the right industries.
The BBC has an excellent tool in which you can search your job title to find out how likely it is to be automated. Check out this link to see jobs that will be automated in the future. Using this tool you will also be able to see jobs least likely to be replaced by robots.
Differences Between Job Automation in Different Countries
The U.K is actually one of the nations that is going to feel the least effects of job automation. Estimates for Germany and the U.S both reveal their face a higher percentage of job automation at 35% and 38% respectively. In studies with more radical results, up to 47% of all U.S jobs are said to be risk from job automation.
There a number of factors for this, primarily differences in distribution of jobs across sectors and differences in task composition within these sectors.
In the U.S for example, the finance and insurance sector is roughly the same size relative to the population of workers as to the U.K’s. However, the U.K’s finance workers are heavily represented within the City of London and as many work within international markets, require higher standards of education and are described as having more complex and difficult roles – therefore, less like susceptible to job automation.
U.S finance professionals are more spread out across the country, more likely to work with local businesses, do not require as much education and have roles described as easier and more routine – thus likely to suffer from job automation.
As for Germany, a significantly higher percentage of workers are found within the manufacturing sector – 11% more than in the U.K. Not only this, but Germany relies less on “Middle Management” in its workplaces and has a higher concentration of workers performing manual and routine tasks – those most likely to be automated.
However, the U.K is not the least automatable country. 21% of Japan's workforce is at risk from job automation, despite higher percentages of the population working in retail and manufacturing. Notably, Japan’s wholesale and retail industry has only 25% of jobs at risk, compared to the U.K’s 44%.
This is mainly in part due to large differences in culture and work ethics, in which the average worker is expected to be much more skilled than their U.K counterparts and spend a lower proportion of their work time in manual and routine tasks, also expected to be regularly performing managerial tasks such as organizing and planning. This applies across every other sector, with each sector being described as much less automatable in comparison to the U.K due the higher demands made of Japanese workers. The only exception was construction, in which 2% more jobs were considered at risk of job automation.
Whilst all these studies base their statistics and findings around what jobs they predict will be ready in the early 2030s, job automation has already happened in some places and is an ongoing process. 44% of firms which have reduced their staff since the financial crisis of 2008 have done so with the via using job automation to fill in the gaps left behind.
Recently, Indian Multinational Infosys confirmed that they have released 11,000 employees through the means of job automation, stating their productivity has increased because of this and with CEO Vishal Sikka commenting that he fully intends to continue down this route and that Infosys will continue to embrace the digital future and the possibilities it brings.
Living in the Future with Job Automation
The idea of human labour being taken by machines isn’t by any means new one. In the early 19th century, many handloom weavers were losing their jobs to mechanisation as the textile industry became more and more reliant on machinery, leading to the Luddite protests in which some workers destroyed the machines that were replacing them.
The transition cannot be described as smooth, but if we are to assume history will repeat itself with this example, then we can expect a heavy initial loss of jobs, but over time for employment to rise back to nearly the same levels and overall levels of wealth increasing.
But assuming the same results would occur when the circumstances are entirely different would be foolish. Robots are beginning to take on more physically and intellectually demanding tasks than ever before and with the advancement of technology faster than ever before - and continuing to accelerate – leads to a very likely possibility that job automation will have a much bigger, long lasting effect on employment – possibly even a permanent one.
The Challenges for Job Automation
Of course, there a still a number of challenges job automation must address before it becomes a widespread practice and not everyone agrees on how best to handle these challenges.
It is important to note that all research conducted into job automation has primarily focused on technological feasibility. There are a number of other economic, legal and regulatory factors in place that might delay the spread of job automation even after the technology exists to make it possible.
Firstly, there’s the price. In almost all cases of technological advancement and invention, there is a period of time between the technology being invented and economically viable to be sold for commercial or corporate use.
When the technology for job automation first becomes reality, the costs of robots and A.I - As well as the costs for their power sources, maintenance and repairs for the robots and security and updates for the A.I) possible updates will be compared to those of human workers, as well as their relative productivity.
The U.K’s rates of employment have been rising steadily over recent years due to negative wage growth (Wage growth which does not meet the rising costs of inflation) and a flexible labour market which has allowed plenty of migration from the EU.
Whilst these conditions are not guaranteed to continue, if they do then many companies will likely see low cost human workers as a more attractive option than to invest in new, expensive technology which has not yet proven itself in a workplace.
However, the economic constraint is one that is certain to become less of an issue over time as the costs of new technologies decreases. Likely only a few companies will be early adopters of this technology, but as they prove effective in these workplaces and their costs fall, other companies will rapidly follow suit.
Legal and regulatory constraints will still apply however, and these may continue to persist even after the economic hurdles have been overcome.
The idea of a “Robot Tax” has been proposed to increase income tax rates for companies who make heavy use of job automation or to assign a “wage” to each robot based on the income it generates, as if it was human and tax this wage like a regular worker. These ideas would discourage companies from rapidly replacing entire workforces with robots and the money from the tax is reinvested into helping those who become victims of job automation.
Critics call this idea an “Innovation Tax” which is designed to hold back progress due to fear of change and also point out that any nation which imposes a robot tax will have a much more difficult time attracting companies to operate their factories there as opposed to countries who do not have such a tax. The “per robot” tax in particular draws criticism from those who argue the definition between machines we already use (Such as computers) and robots is too vague and that machines are all part of a larger technological process – to tax a robot would be to tax a technological process, something which would be unprecedented.
Other proposed solutions are for governments to invest more in vocational education and training. Perhaps the most controversial proposed solution is the introduction of a Universal Basic Income (UBI), a small amount of money that would be given to everyone, regardless of employment status or wealth to ensure even those without work still have the means to survive and to provide the economy with the consumption it needs to survive.
This draws criticism on the grounds of practical affordability, negative impacts on incentive to work and there are simply many who are strongly opposed to the idea of “money for doing nothing” regardless of its intentions or results. UBI is already being trialled in several countries such as Finland, Netherland, select U.S and Canadian states and Brazil. The exact details and methods behind each of these trials differs considerably and as such, the results and data that arise from these trials will likely also differ, but once these trials have concluded and been analyzed we’ll have a more concentrate answer as to the feasibility of UBI.
To Wrap Things up...
Not every study conducted on job automation comes to the same conclusions and every study leaves a lot of room for uncertainty, with the rate of technology’s advancement becoming so fast it is very difficult to predict. OECD’s research came to the conclusion that only 10% of jobs faced a high risk of job automation and that job automation would only take over specific tasks and duties within professions, rarely the professions themselves. Their study shows that job automation would support human workers, not subverting them. It must be noted however, OECD are the outliers in this topic and the only group to come to such a conclusion. The one common verdict shared by all these studies is that to education and training will become more vital to human workers trying to stay one step ahead of robots in the future.
One thing is certain; to succeed in a future where jobs are up for grabs by robots and A.I, Knowledge Is Key.
Studies suggest that Advertising is one of the industries least likely to be affected by job automation soon. Why not read one of our articles on TV advertising and get yourself clued up on this rapidly changing industry?
- Bloomberg, 2017.Job automation risk. Retrieved from bloomberg.com.
- Bank Of England, 2017.Labour’s Share – a speech by Andy Haldane, Chief Economist - Retrieved from the bank of england publications.
- Bank Of England, 2017.Labour’s Share - speech by Andy Haldane - Retrieved from the bank of england publications.
- Bank Of England, 2017.Labour’s Share - Retrieved from the bank of england publications.
- McKinsey & Company, 2017.Harnessing automation for a future that works
- BBC, 2017.Will a robot take your job?
- The Independent, 2017.UK pay growth rate misses expectations as worker real wages turn negative
- The Guardian, 2017.Why robots should be taxed if they take people's jobs
- The Telegraph, 2017.Robots that take people's jobs should pay taxes, says Bill Gatesa>
- Techcrunch, 2017.Is a “robot tax” really an “innovation penalty”?